PROJECT SUMMARY Therapy for estrogen-receptor positive (ER+) breast cancer usually involves resection of the tumor, followed by radiation or chemotherapy, and then long-term treatment with an anti-estrogen therapy such as Tamoxifen. Anti-estrogen therapies target either the production of estrogen, the use of estrogen by the estrogen receptor, or the estrogen receptor itself. As with other targeted therapies, long-term treatment usually results in the remaining cancer cells becoming resistant to the therapy. For the case of ER+ breast cancer, the initial response to anti-estrogen therapy is arrested cell proliferation and reduced cell survival. The continuing stress of therapy causes the remaining cells to switch from a dependence on the estrogen receptor for survival to a dependence on growth factor receptors (GFRs). For cells that successfully make this transition, epigenetic reprogramming may cement the change and lead ultimately to renewed proliferation in spite of therapy. These changes, which occur over the course of months, are initially reversible. To avoid the development of resistance that can result from constant, prolonged anti-estrogen therapy, we propose repeated cycles of an optimized sequence of targeted therapies and rest intervals. The possible targeted therapies include various anti-estrogens, growth-factor receptor inhibitors, histone deacetylases, and DNA methyl transferases. The sequence and timing will be designed to maximize cancer cell death during each cycle, limit the toxicity to normal cells, and return the remaining cells to their original, sensitive state, thereby minimizing the potential for developing resistance. To optimize the sequencing and timing of therapies will require a dynamic mathematical model that captures key cellular adaptations to targeted therapies over time scales of days and months. Techniques from mathematical optimization can then be used to determine optimized therapeutic protocols. To create the dynamic model, we propose a coordinated program of experimentation and mathematical modeling. The experiments and model will consider both short-term biochemical changes in the regulatory network, which occur over the course of days, as well as intermediate- term epigenetic changes, which occur over weeks and months. The experimental work will primarily use well- established ER+ human breast cancer cell lines, with mathematical models and protocols being built for at least two different cell lines to provide some understanding of the impact of genomic heterogeneity. Xenograft models will be used for additional protocol validation in the context of a more realistic tumor microenvironment. Successful completion of this work will provide insights into scheduling therapies for increased effectiveness and avoidance of resistance. Our experiments and models will provide a strong base for future work with primary human tissue to provide clinically actionable results. 1